6 perf-diff - Read perf.data files and display the differential profile
11 'perf diff' [baseline file] [data file1] [[data file2] ... ]
15 This command displays the performance difference amongst two or more perf.data
16 files captured via perf record.
18 If no parameters are passed it will assume perf.data.old and perf.data.
20 The differential profile is displayed only for events matching both
21 specified perf.data files.
23 If no parameters are passed the samples will be sorted by dso and symbol.
24 As the perf.data files could come from different binaries, the symbols addresses
25 could vary. So perf diff is based on the comparison of the files and
32 Dump raw trace in ASCII.
39 Load module symbols. WARNING: use only with -k and LIVE kernel
43 Only consider symbols in these dsos. CSV that understands
44 file://filename entries. This option will affect the percentage
45 of the Baseline/Delta column. See --percentage for more info.
49 Only consider symbols in these comms. CSV that understands
50 file://filename entries. This option will affect the percentage
51 of the Baseline/Delta column. See --percentage for more info.
55 Only consider these symbols. CSV that understands
56 file://filename entries. This option will affect the percentage
57 of the Baseline/Delta column. See --percentage for more info.
61 Sort by key(s): pid, comm, dso, symbol, cpu, parent, srcline.
62 Please see description of --sort in the perf-report man page.
67 Use a special separator character and don't pad with spaces, replacing
68 all occurrences of this separator in symbol names (and other output)
69 with a '.' character, that thus it's the only non valid separator.
73 Be verbose, for instance, show the raw counts in addition to the
78 Do not show any message. (Suppress -v)
82 Don't do ownership validation.
85 Look for files with symbols relative to this directory.
89 Show only items with match in baseline.
93 Differential computation selection - delta, ratio, wdiff, delta-abs
94 (default is delta-abs). Default can be changed using diff.compute
95 config option. See COMPARISON METHODS section for more info.
99 Show period values for both compared hist entries.
103 Show formula for given computation.
107 Specify compute sorting column number. 0 means sorting by baseline
108 overhead and 1 (default) means sorting by computed value of column 1
109 (data from the first file other base baseline). Values more than 1
110 can be used only if enough data files are provided.
111 The default value can be set using the diff.order config option.
114 Determine how to display the overhead percentage of filtered entries.
115 Filters can be applied by --comms, --dsos and/or --symbols options.
117 "relative" means it's relative to filtered entries only so that the
118 sum of shown entries will be always 100%. "absolute" means it retains
119 the original value before and after the filter is applied.
122 Analyze samples within given time window. It supports time
123 percent with multiple time ranges. Time string is 'a%/n,b%/m,...'
124 or 'a%-b%,c%-%d,...'.
128 Select the second 10% time slice to diff:
130 perf diff --time 10%/2
132 Select from 0% to 10% time slice to diff:
134 perf diff --time 0%-10%
136 Select the first and the second 10% time slices to diff:
138 perf diff --time 10%/1,10%/2
140 Select from 0% to 10% and 30% to 40% slices to diff:
142 perf diff --time 0%-10%,30%-40%
144 It also supports analyzing samples within a given time window
145 <start>,<stop>. Times have the format seconds.microseconds. If 'start'
146 is not given (i.e., time string is ',x.y') then analysis starts at
147 the beginning of the file. If stop time is not given (i.e, time
148 string is 'x.y,') then analysis goes to the end of the file. Time string is
149 'a1.b1,c1.d1:a2.b2,c2.d2'. Use ':' to separate timestamps for different
152 For example, we get the timestamp information from 'perf script'.
154 perf script -i perf.data.old
155 mgen 13940 [000] 3946.361400: ...
157 perf script -i perf.data
158 mgen 13940 [000] 3971.150589 ...
160 perf diff --time 3946.361400,:3971.150589,
162 It analyzes the perf.data.old from the timestamp 3946.361400 to
163 the end of perf.data.old and analyzes the perf.data from the
164 timestamp 3971.150589 to the end of perf.data.
166 --cpu:: Only diff samples for the list of CPUs provided. Multiple CPUs can
167 be provided as a comma-separated list with no space: 0,1. Ranges of
168 CPUs are specified with -: 0-2. Default is to report samples on all
172 Only diff samples for given process ID (comma separated list).
175 Only diff samples for given thread ID (comma separated list).
179 The comparison is governed by the baseline file. The baseline perf.data
180 file is iterated for samples. All other perf.data files specified on
181 the command line are searched for the baseline sample pair. If the pair
182 is found, specified computation is made and result is displayed.
184 All samples from non-baseline perf.data files, that do not match any
185 baseline entry, are displayed with empty space within baseline column
186 and possible computation results (delta) in their related column.
188 Example files samples:
189 - file A with samples f1, f2, f3, f4, f6
190 - file B with samples f2, f4, f5
191 - file C with samples f1, f2, f5
194 x - computation takes place for pair
195 b - baseline sample percentage
199 baseline/A compute/B compute/C samples
200 ---------------------------------------
210 baseline/B compute/A compute/C samples
211 ---------------------------------------
221 baseline/C compute/B compute/A samples
222 ---------------------------------------
234 If specified the 'Delta' column is displayed with value 'd' computed as:
236 d = A->period_percent - B->period_percent
239 - A/B being matching hist entry from data/baseline file specified
240 (or perf.data/perf.data.old) respectively.
242 - period_percent being the % of the hist entry period value within
245 - with filtering by -C, -d and/or -S, period_percent might be changed
246 relative to how entries are filtered. Use --percentage=absolute to
247 prevent such fluctuation.
251 Same as 'delta` method, but sort the result with the absolute values.
255 If specified the 'Ratio' column is displayed with value 'r' computed as:
257 r = A->period / B->period
260 - A/B being matching hist entry from data/baseline file specified
261 (or perf.data/perf.data.old) respectively.
263 - period being the hist entry period value
265 wdiff:WEIGHT-B,WEIGHT-A
266 ~~~~~~~~~~~~~~~~~~~~~~~
267 If specified the 'Weighted diff' column is displayed with value 'd' computed as:
269 d = B->period * WEIGHT-A - A->period * WEIGHT-B
271 - A/B being matching hist entry from data/baseline file specified
272 (or perf.data/perf.data.old) respectively.
274 - period being the hist entry period value
276 - WEIGHT-A/WEIGHT-B being user supplied weights in the the '-c' option
277 behind ':' separator like '-c wdiff:1,2'.
278 - WEIGHT-A being the weight of the data file
279 - WEIGHT-B being the weight of the baseline data file
283 linkperf:perf-record[1], linkperf:perf-report[1]